Machine Learning Projects

Welcome to my Machine Learning Portfolio Repository.

This repository contains a growing collection of machine learning projects that solve real-world problems using industry-standard tools, clean code practices, and comprehensive data workflows — from preprocessing and feature engineering to model training, evaluation, and deployment.

---

Repository Structure

ML-Projects/
│
├── sonar-object-classification/           # Sonar signal classification (Mine vs. Rock)
├── Car-Price-Prediction/                  # Regression + Streamlit app for car pricing
├── diabetes-prediction/                   # Disease classification based on health metrics
├── creditcard-fraud-detector/             # Anomaly detection using XGBoost and others
├── laptop-price-prediction/               # Regression + Streamlit app for laptops
├── stock-price-prediction/                # Stock movement classifier
├── loan-status-prediction/                # Predicting loan approvals
└── README.md                              # Main repository documentation

---

Projects Overview

Project	Description	Type	Model(s) Used	Demo / Notebook
Sonar Object Classification	Classify sonar signals as rock or mine	Classification	KNN, SVM, RF, LR, etc.	Notebook
Car Price Prediction	Predict car prices + Streamlit UI	Regression	Linear Regression	App (Streamlit) • Notebook
Diabetes Prediction	Predict diabetes risk using medical features	Classification	SVM, RF, LR, DT	Notebook
Credit Card Fraud Detection	Detect fraudulent transactions	Anomaly Detection	XGBoost, RF, SVM, LR	Notebook
Laptop Price Prediction	Predict laptop prices + Streamlit UI	Regression	RF, XGBoost, Linear, etc.	App (Streamlit)
Stock Price Prediction	Classify next-day stock movement	Classification	SVC, KNN, RF, XGBoost	Notebook
Loan Status Prediction	Predict loan approvals using applicant data	Classification	SVM (linear kernel)	Notebook

---

Completed Projects

🔹 Sonar Object Classification

Binary classification of sonar signals to detect whether the object is a mine or a rock using a range of supervised machine learning models.

• Algorithms: Logistic Regression, SVM, KNN, Decision Tree, Random Forest, Naive Bayes, Gradient Boosting
• Dataset: UCI Sonar Dataset
• Libraries: scikit-learn, pandas, numpy, matplotlib, seaborn
• Highlights:
  • Feature scaling with StandardScaler
  • Principal Component Analysis (PCA) for 2D visualization
  • Comparison of 7 classifiers
  • Best model: K-Nearest Neighbors (90.48% accuracy)
  • Dark-themed visualizations (saved in graphs/ folder)

---

🔹 Car Price Prediction

Predicts used car prices using a Linear Regression model with a Streamlit web interface.

• Algorithm: Linear Regression
• Libraries: scikit-learn, pandas, numpy, streamlit, pickle
• Highlights:
  • Feature engineering and data cleaning
  • Deployment-ready interface for real-time prediction
  • Modular scripts and saved model

---

🔹 Diabetes Prediction

Predicts diabetes likelihood using health-related metrics with multiple classifiers.

• Algorithms: SVM, Logistic Regression, Random Forest, Decision Tree
• Dataset: Pima Indians Diabetes Dataset
• Highlights:
  • Dark-mode visualizations
  • Model comparison and metric evaluation
  • User input prediction via script

---

🔹 Credit Card Fraud Detection

Detects fraudulent transactions using advanced classification models including XGBoost.

• Algorithms: XGBoost, Logistic Regression, SVM, Random Forest, KNN
• Highlights:
  • EDA with DataCmp and visualizations
  • XGBoost achieved 99.96% accuracy
  • Exported model for deployment

---

🔹 Laptop Price Prediction

Predicts laptop prices using regression models and delivers results via a Streamlit app.

• Features: Screen resolution, CPU, GPU, RAM, storage, and more
• Highlights:
  • Rich EDA and feature engineering
  • Deployment-ready with modular code
  • Comparison of regression models

---

🔹 Stock Price Prediction

Predicts next-day stock movement using engineered financial features and ML classifiers.

• Algorithms: SVM, Random Forest, KNN, XGBoost, etc.
• Highlights:
  • Binary classification (Up/Down)
  • Accuracy heatmap comparison
  • GridSearchCV tuning

---

🔹 Loan Status Prediction

Predicts loan approval status based on financial and personal attributes.

• Algorithm: Support Vector Machine (Linear Kernel)
• Highlights:
  • Handling of missing values and categorical encoding
  • Clean visualizations and performance metrics
  • Well-structured scripts and notebook

---

Tech Stack

• Languages: Python 3.7+
• ML Libraries: scikit-learn, xgboost, joblib, pandas, numpy
• Visualization: matplotlib, seaborn, streamlit
• Other Tools: Jupyter Notebook, Git, VSCode, Kaggle, DataCmp

---

About Me

Moustafa Mohamed
Aspiring AI Developer | Focused on Machine Learning, Deep Learning & LLM Engineering

• GitHub: MoustafaMohamed01
• Linkedin: Moustafa Mohamed
• Kaggle: moustafamohamed01
• Portfolio: moustafamohamed

---

⭐ Contribute & Support

If you find this repository helpful, please consider giving it a ⭐.
Contributions, feedback, and ideas are always welcome!

---